2002.-To determine whether reactive oxygen species (ROS) play an essential role in hypoxic pulmonary vasoconstriction (HPV) and the cellular locus of ROS production and action during HPV, we measured internal diameter (ID) at constant transmural pressure, lucigenin-derived chemiluminescence (LDCL), and electron paramagnetic resonance (EPR) spin adduct spectra in small distal porcine pulmonary arteries, and dichlorofluorescein (DCF) fluorescence in myocytes isolated from these arteries. Hypoxia (4% O 2) decreased ID, increased DCF fluorescence, tended to increase LDCL, and in some preparations produced EPR spectra consistent with hydroxyl and alkyl radicals. Superoxide dismutase (SOD, 150 U/ml) or SOD ϩ catalase (CAT, 200 U/ml) did not alter ID during normoxia but reduced or abolished the constriction induced by hypoxia. SOD also blocked HPV in endotheliumdenuded arteries after restoration of the response by exposure to 10 Ϫ10 M endothelin-1. Confocal fluorescence microscopy demonstrated that labeled SOD and CAT entered pulmonary arterial myocytes. SOD, SOD ϩ CAT, and CAT blocked the increase in DCF fluorescence induced by hypoxia, but SOD ϩ CAT and CAT also caused a stable increase in fluorescence during normoxia, suggesting that CAT diminished efflux of DCF from cells or oxidized the dye directly. We conclude that HPV required increased concentrations of ROS produced by and acting on pulmonary arterial smooth muscle rather than endothelium. vascular smooth muscle; endothelium; electron paramagnetic resonance; dichlorofluorescein; lucigenin; superoxide dismutase; catalase REACTIVE OXYGEN SPECIES (ROS) are thought to play important roles in a wide variety of cellular processes (10,22). In the pulmonary circulation, O 2 -dependent changes in ROS concentrations in pulmonary arterial smooth muscle have been proposed by several laboratories to mediate hypoxic pulmonary vasoconstriction (HPV); however, the details of these hypotheses differ greatly. On one hand, hypoxia is thought to reduce production of ROS by the mitochondrial electron transport chain (METC) (3) or a microsomal NADH oxidase (33, 34). On the other hand, hypoxia is thought to increase production of ROS by the METC (53) or a sarcolemmal NADPH oxidase (13,30,59). Proposals for effector pathways are equally varied and include redox-dependent inactivation of sarcolemmal voltagedependent potassium channels and secondary membrane depolarization (57), NADH-dependent enhancement of cyclic ADP ribose production and secondary activation of ryanodine receptors and calcium release from sarcoplasmic reticulum (60), and H 2 O 2 -and catalase-dependent, nitric oxide-independent activation of soluble guanylate cyclase (33,34).None of these proposals includes endothelial cells, which can be a potent source of ROS and other vasoactive factors (7,29), and are thought to play an essential role in HPV (52). In small distal porcine pulmonary arteries, we found that HPV was 1) similar to HPV in isolated or intact lungs in terms of magnitude, O 2 dependence, temporal charact...